NavMarkAR: A Landmark-based AR Wayfinding System for Enhancing Older Adults' Spatial Learning

Wayfinding in complex indoor environments can be challenging for older adults due to natural declines in navigational and spatial-cognition abilities. To assist in improving older adults’ spatial navigation skills, we developed NavMarkAR, an augmented reality (AR) navigation system using smart-glasses to provide landmark-based navigational guidance. This project fills an important gap in design research and practice, as there have been few prior efforts to evaluate the cognitive implications of AR-aided navigational systems. We conducted an initial usability test with 6 participants, followed by design refinement of the prototype and then a more in-depth study with 32 participants, who were asked to use the NavMarkAR system while completing navigation tasks in a university building. We found that participants using NavMarkAR completed the wayfinding tasks faster and developed more accurate cognitive maps. In future research we will study the long-term cognitive skill maintenance effects of using such navigational aids.

TEAM

Zhiwen (Calvin) Qiu, Mojtaba Ashour, Xiaohe Zhou

ADVISOR

Dr. Saleh Kalantari, Dr Abe Davis

LINK

arXiv

PROJECT TYPE

Research, Cornell University

SUBMISSION

Under Review for Advanced Engineering Informatics

Problem Statement

Previous research indicated that AR can significantly reduce time and cognitive load of human wayfinding behaviors (Kim et al., 2015), and well-designed HMD can improve situation awareness in navigation and visual perception tasks (McKendrick et al., 2016). Below are temporary research questions for this study:

RQ1. How will a landmark-based AR navigation system affect wayfinding performance and the development of cognitive maps in an older adult population?

RQ2. What are the preferences of older adults regarding the multimodal interactions and user experience in landmark-based AR navigation systems?

Prototype Design

Thee AR navigation system, developed for HoloLens 2, integrates a series of advanced techniques for optimal user navigation. It first employs spatial mapping to scan the 3D environment using the Scene Understanding SDK, creating a virtual replica of the real world. The system then utilizes the HoloLens camera and other sensors to accurately localize users within this environment. To provide real-time pathfinding, the system applies the A* algorithm, renowned for its efficiency and accuracy. Lastly, we enrich the user interface with virtual instructions, augmented landmarks, and other pertinent cues to guide the user, thus providing a comprehensive, intuitive, and interactive navigation experience.